ON THE HORIZONTAL MOON. 15$ 



objects of known magnitudes, has thejgreateft (hare in pro- 

 ducing tiie iliufion by which it appears largeft in tl.e horizon, 

 I would by no means be underllood to reject the principal 

 circumiianoes, hmg fince noticed as contributing towards the 

 lame effect. Different circumilances will neceflarily have 

 more or lefs influence with different perfons, according to 

 the tenor of their previoufly acquired habits and observations. 

 I remain, Sir, 



Your obedient fervant, 



C. H. 

 Tavi flock Place. Feb. 15. 



ANNOTATION. W. N. 



THIS very (hiking and perfpicuous explanation of the Experiment 

 enlargement of the moon near the horizon, will bring to the Wlt ^ Hadley's 

 reader's recollection a great number of fads, where the ab- which the fun 

 folute magnitudes of objects, of which the diftances are im- and mo011 appear 

 perfectly known, are erroneoufly eflimated by comparifon nitudeT^ccord- 

 with other objects lying in the fame direction. Navigators \ n & t0 the ob- 

 muft have been often flruck with the extreme minutenefs of their imaee 1C 

 the image of the fun or moon, when brought down by piojefled. 

 Hadley's quadrant fo as to be feen projected upon the near 

 objects below the horizon. Since the perufal of this com- 

 munication, I have repeated the experiment on thofe lumi- 

 naries. If the image of the moon be removed a certain num- 

 ber of degrees from the direct ray, it is eafy, by altering the 

 pofition of the quadrant, to obferve that image in the horizon 

 to the right or left, or upon the pavement before us, or at a 

 great elevation in the heavens. In thefe cafes its apparent 

 magnitude, while the angular fize and brightnefs remain un- 

 altered, is found to be frhall at the great elevation, large in 

 the horizon, and mod fo when projected upon diitant objects, 

 and it becomes a minute fpeck when feen depreffed among 

 the objects clofe to the obferver. The fame experiment may 

 be made with a pane of glafs, but lefs objectionably, becauie 

 it might be urged that the quantity of illumination is different 

 according (o* the obliquity of reflection. 



A Math- 



